Does hometown advantage matter? The case of institutional blockholder monitoring on earnings management in Korea

Abstract

This study uses a large sample of Korean firms to examine institutional blockholders’ influence on corporate earnings management. The Korean market is an interesting setting to study institutional monitoring because it is dominated by chaebols, which are characterized by ineffective internal governance mechanisms. We find evidence that institutional blockholders deter opportunistic financial reporting, and such activities are the most evident among domestic institutional blockholders. Foreign institutional blockholders, especially those with short-term investment objectives, lead to more discretionary reporting. Our findings are robust to various empirical models and endogeneity checks. Moreover, this study offers new insight into the proximity advantage of domestic institutional investors with regard to monitoring information. It also highlights institutional investors’ roles in shaping corporate governance in emerging markets, in which sound corporate governance has become increasingly important for firms’ long-term sustainability. One important policy implication is that regulatory authorities should aim to promote institutional investors’ participation in these markets.

Résumé

Cette étude utilise un large échantillon de firmes coréennes pour étudier l’influence des actionnaires institutionnels dominants sur la gestion des résultats des sociétés. Le marché coréen constitue un contexte intéressant pour étudier le contrôle institutionnel parce qu’il est dominé par des chaebols qui sont caractérisés par des mécanismes de gouvernance interne inefficients. Nous constatons que les actionnaires institutionnels dominants détériorent le reporting financier opportuniste, et que ces activités sont plus visibles parmi les actionnaires institutionnels dominants domestiques. Les actionnaires institutionnels dominants étrangers, plus particulièrement ceux ayant des objectifs d’investissement à court terme, donnent lieu à un reporting plus discrétionnaire. Nos résultats sont robustes face à plusieurs modèles empiriques et à des vérifications d’endogénéité. Par ailleurs, cette étude offre un nouvel éclairage sur l’avantage de proximité des investisseurs institutionnels domestiques concernant le contrôle de l’information. Elle met aussi en relief les rôles des investisseurs institutionnels dans la structuration de la gouvernance d’entreprise dans les marchés émergents où une gouvernance d’entreprise solide est devenue plus importante pour la pérennité des firmes sur le long terme. Une implication politique importante est que les autorités de régulation devraient s’efforcer de promouvoir la participation des investisseurs institutionnels sur ces marchés.

Resumen

Este estudio usa una amplia muestra de empresas coreanas para examinar la influencia de los bloqueadores institucionales en la gestión de las ganancias corporativas. El mercado coreano es un escenario interesante para estudiar el monitoreo institucional porque está dominado por chaebols, los cuales están caracterizados por mecanismos ineficaces de gobernanza interna. Encontramos evidencias que los bloqueadores institucionales hacen disuadir los reportes financieros oportunistas, y estas actividades son las más evidentes entre los bloqueadores institucionales nacionales. Los bloqueadores institucionales extranjeros, especialmente aquellos con objetivos de inversión a corto plazo, conducen a reportes discrecionales. Nuestros hallazgos son robustos a varios modelos empíricos y controles de endogeneidad. Además, este estudio ofrece un nuevo entendimiento de la ventaja de proximidad de los inversionistas domésticos institucionales con respecto al monitoreo de información. También resalta los roles de los inversionistas locales en la configuración de la gobernanza corporativa en mercados emergentes, en los que el buen gobierno corporativo se ha vuelto cada más importante para la sostenibilidad de largo plazo de las empresas. Una importante implicación para la política es que las autoridades regulatorias deben tener como objetivo promover la participación institucional de los inversionistas en estos mercados.

Resumo

Este estudo usa uma grande amostra de empresas coreanas para examinar a influência de blockholders institucionais sobre o gerenciamento de ganhos corporativos. O mercado coreano é um cenário interessante para estudar o monitoramento institucional porque é dominado por chaebols, caracterizados por mecanismos de governança interna ineficazes. Encontramos evidências de que os blockholders institucionais evitam a divulgação oportunista de relatórios financeiros, e que essas atividades são mais evidentes entre os blockholders institucionais nacionais. Os blockholders institucionais estrangeiros, especialmente aqueles com objetivos de investimento de curto prazo, levam a relatórios mais discricionários. Nossas descobertas são robustas para vários modelos empíricos e verificações de endogeneidade. Além disso, este estudo oferece uma nova visão da vantagem de proximidade dos investidores institucionais nacionais em relação à monitoração de informações. Também destaca os papéis dos investidores institucionais na definição da governança corporativa nos mercados emergentes, em que uma sólida governança corporativa tornou-se cada vez mais importante para a sustentabilidade a longo prazo das empresas. Uma importante implicação política é que as autoridades reguladoras devem procurar promover a participação dos investidores institucionais nesses mercados.

概要

本研究用韩国公司的大样本来研究机构大股东对公司收益管理的影响。韩国市场由于被特点是低效的内部治理机制的财团所主导,因而是研究机构监控的一个有趣环境。我们发现的证据表明,机构大股东阻止投机的财务报告,此类活动在国内机构大股东中最为明显。外国机构大股东,特别是那些投资目标短期的,会导致更多任意决定的报告。我们的发现对各种实证模型和内生检查是稳准的。此外,这项研究在监控信息方面对国内机构投资者的接近优势有新见地。这也凸显了在新兴市场公司治理塑造中机构投资者的角色,其中健全的公司治理对公司长期可持续发展日益重要。一个重要的政策启示是,监管部门应着眼于促进机构投资者在这些市场中的参与。

Appendices

Appendix 1

See Table A1.

Table A1 Sample selection process

Appendix 2

See Table A2.

Table A2 Definitions of the control variables

Appendix 3

Reverse Causality Checks

We perform multiple tests in this Appendix to check whether our results are driven by the reverse causality between earnings quality and institutional blockholdings.

We start by examining the relationship between lagged earnings management and institutional ownership. If our results arise from domestic institutional blockholders’ ability to pick firms with high-quality earnings, we should expect a significant relationship between the two. Table A3.1 shows that, generally, the lagged D_ADAs are not significantly related to institutional blockholdings except for the three-year lagged D_ADA with BLOCK_IO and BLOCK_IO_F. More importantly, we do not find any significant relationship between lagged D_ADA and BLOCK_IO_D, implying that our main finding is not affected by reverse causality.

Table A3.1 Preference of institutional blockholders for earnings management

Additionally, we consider the change in D_ADA (ΔD_ADA) to be our response variable. While the lead-lag setup used in our main analysis helps alleviate endogeneity concerns, it may not work as intended if the level of earnings management persists over time. We address this issue by following Chung et al. (2015) and use the change in earnings quality as the dependent variable. Table A3.2 reports the results of using ΔD_ADA as the dependent variable; we find that our main results remain unchanged.

Table A3.2 Influence of institutional blockholdings on change in D_ADA

Further, we use the 2SLS method to check our results’ robustness. Specifically, we use the first-stage regression to model institutional preferences following the specifications of Kim et al. (2016a). Specifically, we consider two instruments for institutional blockholding – DIVR and TVOL – with DIVR as the common stock cash dividend divided by the market value of equity for firm i in year t, and TVOL as the trading volume divided by the common shares outstanding for firm i in year t. Bennett, Sias, and Starks (2003) note that aggregate institutional investors prefer to invest in dividend-paying firms. Kim et al. (2016a) document that both domestic and foreign institutional ownerships are positively related to high dividend-paying firms, while such a relationship is more pronounced with foreign institutional investors, as in Ferreira and Matos (2008). Additionally, Brennan and Cao (1997) and Grinblatt and Keloharju (2000) find that foreign institutional investors favor holding more liquid stocks than their domestic counterparts. The first two columns in Table A3.3 report the estimation results from the first-stage regression with domestic and foreign institutional blockholdings as the dependent variables, respectively. The third column reports the estimation results from the second-stage regression, in which we regress earnings management on the instrumented domestic (P_BLOCK_IO_D) and foreign (P_BLOCK_IO_F) institutional blockholdings. Consistent with our main finding and the “hometown advantage” hypothesis, we find that domestic institutional blockholdings are negatively related to earnings management, while foreign institutional blockholdings are positively related.

Table A3.3 Influence of institutional blockholdings on earnings management: 2SLS and GMM estimation approach

Finally, Wintoki, Linck, and Netter (2012) propose the dynamic panel GMM approach to address endogeneity problems in corporate finance, in which researchers often face a lack of natural experiments and exogenous instruments. We follow a similar methodology as in Kim et al. (2016a) for our results to be comparable, and control for one-, two-, and three-year lagged D_ADAs as the instrumental variables in the GMM estimation. The last two columns of Table A3.3 indicate that domestic institutional blockholdings are significantly and negatively related to earnings management, while foreign institutional blockholdings are insignificant. Additionally, the Hansen test for over-identifying restrictions (J-statistic) reveals that we cannot reject that our choice of GMM instruments is valid.

Appendix 4

See Table A4.

Table A4 Influence of institutional blockholdings on signed ADA